A few comments:
You are measuring bass response at just your mix position? That's great if you only mix in the room, but if you record at various spots, you should test different spots too. Your unexplained result on opening a window could be because a node moved as a result (just a guess though). On the other hand, the corner door open would act like a very good corner absorber.
I think a sine wave RMS would be -6dB, not 0dB. I'll have to check that.
The recommended monitor placement is for proper imaging, not avoidance of bass nodes. Imaging is more a high-frequency phenomenon. Treat your room completely, and place the monitors as recommended (although 3 ft is tighter than I like mine).
Finally, don't use an EQ curve to flatten your room response. Add more absorption to the room.
Edit: Oh yeah! The obvious--you have to adjust for the response of your monitors--they are going to naturally fall off at lower frequencies--so you shouldn't expect to be able to achieve a flat curve. Concentrate on getting the peaks flattened, but not the overall level up.
How long a delay did you set between frequencies? It takes a couple of msec for the bass waves to travel around the room to create nodes.
